Pierce Krouse wrote:
>>The problem with contoured formats
is that they pretty much enforce a given layer thickness when it comes
time to build a prototype.
Without starting a flame war as to which is the better format, I'm
sure others in the group will be happy to elaborate further :-) <<
As always, I am happy to elaborate:
1) Tesselated formats are not so good when we are dealing with trim curves on
surface models. In fact many surface modeling packages cannot properly tesselate
many of the constructions that they can render. I am not passing judgement on
the idea, just on the current state of the art. Brock Rooney's IGES to STL does
a very good job in this regard. Unfortunately, most surface modeling vendors are
not that smart.
2) Contour data is easier to edit, on a per layer basis. Modifications that span
several layers are a royal pain.
3) Contour data provides better fidelity to the original surface, provided it is
done right. If the original surface is tesselated to produce the contour, then
you have gained nothing.
4) Contour data does not use surface normal information. A definate advantage
given the number of files out there with messed up normals.
5) Contour data cannot be easily visualized.
6) Contour data cannot be re-oriented without introducing approximation errors
of the order of the slice thickness
7) As long as the contour data file has slice data at fine enough z levels, (say
every 0.0005"), there is no need to worry about using different slice thickness
for the build. A negative is that the file is unduly large.
8) Processing contour data from surface models can be very time consuming.
Tesselation is much faster.
9) Contour generation is more easily adaptable to hardware acceleration.
10) Support structures are difficult to generate and edit with contour data.
(Walking to the closet looking for my flame retardent suit...)
Peter H. Gien
POGO International, Inc.
This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:37:13 EEST